rafaelvalle / asrgen Goto Github PK
View Code? Open in Web Editor NEWAttacking Speaker Recognition with Deep Generative Models
Home Page: https://arxiv.org/pdf/1801.02384.pdf
Attacking Speaker Recognition with Deep Generative Models
Home Page: https://arxiv.org/pdf/1801.02384.pdf
Thank you for your contribution, I have some doubts in the experiment, I hope you can answer.
First question:
In gan_synthesis.ipynb,
audio = load_wav_to_torch('data_16khz/zcathy/cathy.wav', SAMPLING_RATE)
audio /= MAX_WAV_VALUE
audio = audio[None, :]
reference_mel = taco_stft.mel_spectrogram(audio)[0]
print(reference_mel.min(), reference_mel.max())
mel -= mel.min()
mel = mel / mel.max()
mel = mel * reference_mel.max()
print(mel.min(), mel.max())**
Is mel = mel * reference_mel.max() the matching of the generated fake audio with the real audio?
I don't quite understand how to use the trained G_NET to generate the voiceprint audio that matches the target.
Second question:
Is gan_attack.ipynb a target attack?
The target ID you set is 0. Can this be modified and replaced with another ID?
Looking forward to your reply!
‘python python gan_train.py ’
An error has occurred
File "", line 1, in
runfile('D:/Documents/paper/asrgen-master/asrgen-master/gan_train.py', wdir='D:/Documents/paper/asrgen-master/asrgen-master')
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 827, in runfile
execfile(filename, namespace)
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\spyder_kernels\customize\spydercustomize.py", line 110, in execfile
exec(compile(f.read(), filename, 'exec'), namespace)
File "D:/Documents/paper/asrgen-master/asrgen-master/gan_train.py", line 140, in
logger.log_validation(real_data_spk, real_data_spk+reg_noise,fake_data, iteration)
File "D:\Documents\paper\asrgen-master\asrgen-master\logger.py", line 33, in log_validation
iteration)
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\tensorboardX\writer.py", line 548, in add_image
image(tag, img_tensor, dataformats=dataformats), global_step, walltime)
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\tensorboardX\summary.py", line 216, in image
image = make_image(tensor, rescale=rescale)
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\tensorboardX\summary.py", line 256, in make_image
image = Image.fromarray(tensor)
File "D:\Program Files\Anconda3\envs\tensorflow\lib\site-packages\PIL\Image.py", line 2492, in fromarray
raise TypeError("Cannot handle this data type")
TypeError: Cannot handle this data type
How can I fix it ?
Sorry,I can't understand why the 82nd (or 94th)line in gan_train.py uses BATCH_SIZE instead of SAMPLE_SIZE. Because in my view,when the 62nd line in gan_train.py uses SAMPLE_SIZE,we actually get train_generator with 2*SAMPLE_SIZE.Then the boundary in the 82nd line should be SAMPLE_SIZE.Where is wrong?I'm not good at it.Thanks sincerely.
Can you tell me which datasets should be used to train the speaker recognition system while there is a folder named 'data_16khz', and I doubt it contains 100 speakers of 2004 NIST and 1 speaker of 2013 Blizzard and I don't know is this enough? or maybe I don't get the complete audios of each speaker which makes the accuracy of sr is extremely low? Where can I get complete dataset for the speaker recognition system.Thanks.
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.